Obstructive sleep apnea (OSA) is a serious sleep disorder affecting 2-9% of the US population. It is caused by recurrent obstruction of the upper airway (velopharynx and oropharynx) during sleep and produces daytime sleepiness, and increases cardiovascular risk and mortality. Treatment with continuous positive airway pressure (CPAP) is effective and reduces behavioral and cardiovascular risk, but 40% of patients with moderate to severe disease cannot or will not tolerate this first line therapy, and alternatives no very predictable as long-term treatment. A barrier to testing neurostimulation approaches for OSA is the lack of a reliable tool for development and testing of technology, effectiveness, and off-target effects. The proposal is to develop and verify clinical correlates of OSA in a rabbit model of OSA, based on it having general anatomic similarity to the human upper airway, and its size, cost, and temperament. We will produce recurrent obstruction during sleep by partial nasopharyngeal obstruction, airway crowding produced by injection of a silicone filler in the base of the tongue, and verify the endpoints present in human OSA, including sympathetic excitation (increasing heart rate and blood pressure) and sleep instability. We will characterize site(s) of obstruction and the upstream-pressure-flow behavior of the airway. The model will be tested by unilateral hypoglossal nerve stimulation (HNS), and compared to carotid sinus nerve stimulation (CNS) which has an ability to activate and coordinate bilateral upper airway muscle activation through brainstem mechanisms. Aim 1 is to develop, verify, and examine the production of upper airway obstruction acutely under anesthesia and Aim 2 is to record selected consequences during sleep and its stages, intermediate endpoints in the pathology of human OSA. In addition, electrodes will survey cortical state-related evoked potentials and respiratory muscle activation, and blood pressure and heart rate variability will assay autonomic efferent effects. We will mitigate OSA by HNS and CNS. Cuff electrodes will provide selective stimulation. Stimulus parameters will initially be classically-based, and move towards non-traditional paradigms using varying frequency and amplitude, to activate appropriate efferent vs. the afferent fibers. The deliverables in Aim 1 are to demonstrate feasibility and functions, using stimulation approaches to alter upper airway stiffness and resistance and examine respiratory control during drug-induced surgical anesthesia. In Aim 2, we verify the stability and fidelity of the model to human OSA, monitoring sleep (in)stability and autonomic outcomes. We will use HNS to immediately reverse OSA, and study its effects on on-target velopharyngeal and oropharyngeal sites for therapeutic intent, mitigation of sympathetic excitation, and off-target effects on the sensory or motor cortex and autonomic reflex actions. This application creates a tool where scientists in respiratory control, upper airway physiology, and biomedical engineering can address model neurotherapeutic efficacy and side effects as treatment for a common sleep disorder.